Potentiometer 5 or 6 detents

[Travis Covington]

My plan was to mount it on the RC.

[Adam Shapiro]

Originally Posted by Travis Covington

My plan was to mount it on the RC.

The RC has the same thing going for it. The only thing I would be concerned with however is that using a rig like this might be a bit cumbersome. Personally I'd use a 100k ohm potentiometer, rather than a rotary switch, and just set specific ranges as certain programs. This could also be reprogrammed to allow for smaller ranges -> more programs. That's up to you though. The math/physics behind the voltage divider idea would stay the same though, you'd just hook it to an analog input on the RC rather than one on the OI.

[Alan Anderson]

Originally Posted by Adam Shapiro

The idea here isn't as much the actual resistance as the concept...

If the switch is going on the OI, the actual resistance is rather important. The OI analog inputs want to see a variable resistance between ~2k and 100k to give the full range of values.

For a six-position rotary switch, I'd use six 15k resistors wired like this:

+5 <----/\/\/--*--/\/\/--*--/\/\/--*--/\/\/--*--/\/\/--*--/\/\/--*
                         ^
 OI                      |
analog input <-----------'

The * characters represent the six positions of the rotary switch, and the ^ is the common connection.

What you'll get from this is the equivalent of a joystick with six discrete resistance values: 15k, 30k, 45k, 60k, 75k, and 90k. You should use a dashboard program to determine what the actual numbers end up being, and choose appropriate ranges in the software to decide which of the six positions is chosen.

[Adam Shapiro]

Originally Posted by Alan Anderson

If the switch is going on the OI, the actual resistance is rather important. The OI analog inputs want to see a variable resistance between ~2k and 100k to give the full range of values.

I actually never knew that, though in retrospect I suppose it makes sense. Thanks for the clarification!

[Tom Bottiglieri]

I see a business opportunity for Andy Mark/IFI! Latching potentiometers/ prewired variable resistance rotary switches!

Travis,

I thoroughly googled for a pre-made product that would suit your needs but came up with nothing. The best solution in this thread is Alan Anderson's, which is exactly the method I was going to post before I saw his.

I'd check to see if any of the students on your team are taking/have taken electronics classes at school. If so, you can just buy a rotary switch and some 15 Ohm resistors (if you need more inputs, decrease the Ohm rating of the resistor. What you want = 100 / Number Of Switch Settings + 1. Get all of this stuff at RadioShack) and give it to the student to wire up. Anyone with basic knowledge in wiring should be able to tackle this circuit.

[Joe Ross]

Originally Posted by Tom Bottiglieri

I thoroughly googled for a pre-made product that would suit your needs but came up with nothing. The best solution in this thread is Alan Anderson's, which is exactly the method I was going to post before I saw his.

I found the Bourns 3680 series. However, it is very expensive. http://www.mouser.com/catalog/625/505.pdf

[sanddrag]

In past years we had like three two position switches and we would have a setting like (up down up) or (off off on) for the switches before each match depending on what autonomous mode we were going to run. But such things could get confusing in the rush of competition, and I see your desire for the rotary type control.

[Travis Covington]

The potentiometer I found on digikey seems like it will work, but as Joe told me, is an SMD style and would require some delicate soldering. I think I can handle the soldering, but is there any reason not to use this pot?

http://dkc3.digikey.com/PDF/T061/SectF.pdf

Page 74, 16mm square rotary potentiometer with detents.

Also, I think I can handle the rotary switch and 15k resistor method as well. I think mounting the rotary switch and a knob will be easier than the pot with detents and no shaft with a make shift shaft and knob.

Edit: How would that diagram differ if I put the rotary switch/resistor setup on the RC?

[Alan Anderson]

Originally Posted by Travis Covington

...How would that diagram differ if I put the rotary switch/resistor setup on the RC?

Just add another resistor to ground on the far right, just like the one to +5 on the left. (On the RC, you could actually make the +5 and ground connections directly without resistors, but I'm used to being able to detect short circuits as "impossible" input values.)

[Al Skierkiewicz]

There are pots that exist that have detents built in but there are generally used for volume controls (are log taper, 10K) and have 10-12 detent positions. A better solution is to pick up a six position switch at Radio Shack.
http://www.radioshack.com/search/ind...ar y%20switch

This switch is easily modified to become a twelve position switch by bending one of the stopper tabs out of the way and cutting one of the contact springs. Wire 9.1K ohm resistors from each terminal to the one next to it to form a series string. When you are done, you have a stepped 100K pot. Wire as you would any pot to an analog input on the OI and then set your programmers to work. RS has some nice shiny knobs as well. If you would like to see one in action, stop by our pit in Atlanta.

[Travis Covington]

Are you saying to use the 9.1k resistors if I wanted to have 12 positions? Would I still use the 15k resistors if I were to only use the 6 positions?

[Tom Bottiglieri]

Originally Posted by Travis Covington

Are you saying to use the 9.1k resistors if I wanted to have 12 positions? Would I still use the 15k resistors if I were to only use the 6 positions?

You can use either. If you use the 6 position switch, 15k resistors will give your programmers a little more leeway to distinguish one "step" from another.

[Alan Anderson]

Originally Posted by Travis Covington

Are you saying to use the 9.1k resistors if I wanted to have 12 positions? Would I still use the 15k resistors if I were to only use the 6 positions?

For an OI-connected selector, you definitely want the maximum resistance to be just under 100k, and the minimum resistance should probably be at least 2k to keep from overloading the +5 source. That's why I suggested 15k for six positions. 9.1k will be fine for 11 steps, but I think it's a little too high for comfort if you want 12.

The total series value of an RC-connected "stepped potentiometer" is not that critical, but 100k is still a good target.

[Gary Bonner]

Here is a photo of our switch from last year, which was connected to the RC. It’s a six position switch with five 22k resistors (100k / (# positions - 1) = 20k, 22k was the closest we could find at Radio Shack). The switch positions were then 0k, 22k, 44k, 66k, 88k, & 110k. This would give analog values of about 0, 205, 410, 615, 820, and 1024. In the autonomous code, we looked for a range of 205, centered on each of those values.

void User_Autonomous_Code(void)
{
		.
		.
		.

	switch (((Get_Analog_Value( rc_ana_in01 )+100)/205)+1)	//read autonomous switch
	{  
		case 1:	//switch position 1 
	
			while (autonomous_mode)   /* DO NOT CHANGE! */
			{
				if (statusflag.NEW_SPI_DATA)      /* 26.2ms loop area */
				{
					Getdata(&rxdata);   /* DO NOT DELETE, or you will be stuck here forever! */

					/* Add your own autonomous code here. */
					
					Putdata(&txdata);   /* DO NOT DELETE, or you will get no PWM outputs! */
				}
			}
			break; //end of case 1
	
		case 2:	//switch position 2
	
			.
			.
			.

		default:
		{
			// something's wrong - do nothing
		}
	}  //switch
}  //User_Autonomous_Code

[kaszeta]

Originally Posted by Adam Shapiro

The RC has the same thing going for it. The only thing I would be concerned with however is that using a rig like this might be a bit cumbersome. Personally I'd use a 100k ohm potentiometer, rather than a rotary switch, and just set specific ranges as certain programs. This could also be reprogrammed to allow for smaller ranges -> more programs. That's up to you though. The math/physics behind the voltage divider idea would stay the same though, you'd just hook it to an analog input on the RC rather than one on the OI.

That's what Team 95 has done several times. We just had a pot on the robot, no detents, and ranges marked by the knob corresponding to each strategy. With the little knob we used, this allowed us to do up to 12 positions[1] without worrying about ambiguity, although after BAE we realized that we were only really using 4 of these positions, so we reprogrammed the switches.

I think next year I'll encourage the team to make something for the OI, however.


[1] Which meant that the switch went from 0 to 11. One team member asked "can't it just to to 10 instead." Of course, I had to answer "no, these go to 11." Only one person got the joke.